Digital

Communications
ECE-III-II Sem
R15
 Communication
• Main purpose of communication is to transfer information from a
source to a recipient via a channel or medium.

• Basic block diagram of a communication system:

Source Transmitter Channel Receiver

Recipient
 Brief Description
• Source: analog or digital
• Transmitter: transducer, amplifier, modulator, oscillator, power amp.,
antenna
• Channel: e.g. cable, optical fibre, free space
• Receiver: antenna, amplifier, demodulator, oscillator, power amplifier,
transducer
• Recipient: e.g. person, (loud) speaker, computer
• Types of information
Voice, data, video, music, email etc.

• Types of communication systems

Public Switched Telephone Network (voice,fax,modem)
Satellite systems
Radio,TV broadcasting
Cellular phones
Computer networks (LANs, WANs, WLANs)
 Information Representation
• Communication system converts information into electrical
electromagnetic/optical signals appropriate for the transmission medium.
• Analog systems convert analog message into signals that can propagate
through the channel.

• Digital systems convert bits(digits, symbols) into signals

– Computers naturally generate information as characters/bits

– Most information can be converted into bits
– Analog signals converted to bits by sampling and quantizing (A/D
conversion)
 Why digital?
 Digital techniques need to distinguish between discrete symbols allowing
regeneration versus amplification

 Good processing techniques are available for digital signals, such as

medium.
 Data compression (or source coding)
 Error Correction (or channel coding)(A/D conversion)
 Equalization
 Security

 Easy to mix signals and data using digital techniques

• Basic Digital Communication Transformations
– Formatting/Source Coding
– Transforms source info into digital symbols (digitization)
– Selects compatible waveforms (matching function)
– Introduces redundancy which facilitates accurate decoding
despite errors
• It is essential for reliable communication
– Modulation/Demodulation
– Modulation is the process of modifying the info signal to
facilitate transmission
– Demodulation reverses the process of modulation. It involves
the detection and retrieval of the info signal
• Types
• Coherent: Requires a reference info for detection
• Noncoherent: Does not require reference phase information
 Basic Digital Communication Transformations
– Coding/Decoding
Translating info bits to transmitter data symbols
Techniques used to enhance info signal so that they are less
vulnerable to channel impairment (e.g. noise, fading, jamming,
interference)
• Two Categories
– Waveform Coding
• Produces new waveforms with better performance
– Structured Sequences
• Involves the use of redundant bits to determine the
occurrence of error (and sometimes correct it)
– Multiplexing/Multiple Access Is synonymous with resource
sharing with other users
– Frequency Division Multiplexing/Multiple Access (FDM/FDMA
 Main Points
• Transmitters modulate analog messages or bits in case of a DCS for
transmission over a channel.
• Receivers recreate signals or bits from received signal (mitigate
channel effects)
• Performance metric for analog systems is fidelity, for digital it is the bit
rate and error probability.
 Why Digital Communications?
– Easy to regenerate the distorted signal
– Regenerative repeaters along the transmission path can detect a
digital signal and retransmit a new, clean (noise free) signal
– These repeaters prevent accumulation of noise along the path
• This is not possible with analog communication systems
– Two-state signal representation
• The input to a digital system is in the form of a sequence
of bits (binary or M_ary)
– Immunity to distortion and interference
– Digital communication is rugged in the sense that it is more
immune to channel noise and distortion
 Why Digital Communications?
– Hardware is more flexible
– Digital hardware implementation is flexible and permits
the use of microprocessors, mini-processors, digital
switching and VLSI
• Shorter design and production cycle
– Low cost
• The use of LSI and VLSI in the design of components and
systems have resulted in lower cost
– Easier and more efficient to multiplex several digital signals
– Digital multiplexing techniques – Time & Code Division
Multiple Access - are easier to implement than analog
techniques such as Frequency Division Multiple Access
 Why Digital Communications?
– Can combine different signal types – data, voice, text, etc.
– Data communication in computers is digital in nature whereas
voice communication between people is analog in nature
– The two types of communication are difficult to combine over
the same medium in the analog domain.
• Using digital techniques, it is possible to combine both
format for transmission through a common medium
• Encryption and privacy techniques are easier to
implement
– Better overall performance
– Digital communication is inherently more efficient than analog
in realizing the exchange of SNR for bandwidth
– Digital signals can be coded to yield extremely low rates and
high fidelity as well as privacy
 Why Digital Communications?
Disadvantages
• Requires reliable “synchronization”
• Requires A/D conversions at high rate
• Requires larger bandwidth
• Nongraceful degradation
• Performance Criteria
• Probability of error or Bit Error Rate
 Goals in Communication System Design
• To maximize transmission rate, R
• To maximize system utilization, U
• To minimize bit error rate, Pe
• To minimize required systems bandwidth, W
• To minimize system complexity, Cx
• To minimize required power, Eb/No
 Digital Signal Nomenclature

• Information Source
– Discrete output values e.g. Keyboard
– Analog signal source e.g. output of a microphone
• Character
– Member of an alphanumeric/symbol (A to Z, 0 to 9)
– Characters can be mapped into a sequence of binary digits using
one of the standardized codes such as
• ASCII: American Standard Code for Information Interchange
• EBCDIC: Extended Binary Coded Decimal Interchange Code
 Digital Signal Nomenclature

• Digital Message
– Messages constructed from a finite number of symbols; e.g., printed language
consists of 26 letters, 10 numbers, “space” and several punctuation marks.
Hence a text is a digital message constructed from about 50 symbols
– Morse-coded telegraph message is a digital message constructed from two
symbols “Mark” and “Space”
• M - ary
– A digital message constructed with M symbols
• Digital Waveform
– Current or voltage waveform that represents a digital symbol
• Bit Rate
– Actual rate at which information is transmitted per second
 Digital Signal Nomenclature

• Baud Rate
– Refers to the rate at which the signaling elements are
transmitted, i.e. number of signaling elements per second.

• Bit Error Rate

– The probability that one of the bits is in error or simply the
probability of error
 Digital Pulse Modulation
• The process of Sampling which we have already
discussed in initial slides is also adopted in Digital
pulse modulation.
• It is mainly of two types:
Pulse Code Modulation(PCM)
Delta Modulation(DM)
Pulse Code Modulation(PCM)
 Pulse-Code Modulation (PCM) is the most
commonly used digital modulation scheme
 In PCM, the available range of signal voltages is
divided into levels and each is assigned a binary
number
 Each sample is represented by a binary number
and transmitted serially
 The number of levels available depends upon the
number of bits used to express the sample value
 The number of levels is given by: N = 2m
22
22
Pulse Code Modulation(PCM)
 PCM consists of three steps to digitize
an analog signal:
1. Sampling
2. Quantization
3. Binary encoding
 Before we sample, we have to filter the
signal to limit the maximum frequency of
the signal .Filtering should ensure that
we do not distort the signal, ie remove
high frequency components that affect
the signal shape.
23
23
Pulse Code Modulation(PCM)

The basic elements of a PCM system. 24

24
 Pulse Code Modulation(PCM)
Analog to digital converter employs two
techniques:
1. Sampling: The process of generating pulses of
zero width and of amplitude equal to the
instantaneous amplitude of the analog signal.
The no. of pulses per second is called “sampling
rate”.

25
25
Pulse Code Modulation(PCM)
 2. Quantization: The process of dividing
the maximum value of the analog signal
into a fixed no. of levels in order to
convert the PAM into a Binary Code.
The levels obtained are called
“quanization levels”.

26
26
 V Sampling,
o Quantization and
l Coding
t
a
g
e Time
7 111
L 110
B
6 C
e 5 101 i
v n o
4 100
e 3 011 a d
l 2 010 r e
s 1 001 s
Time 000
y
0
V
o 01010111011111010101 0
l
t
a
g
e 27
Time 27
Pulse Code Modulation(PCM)

28
28
 Quantization
• By quantizing the PAM pulse, original signal is only
approximated
• The process of converting analog signals to PCM is
called quantizing
• Since the original signal can have an infinite number
of signal levels, the quantizing process will produce
errors called quantizing errors or quantizing noise

29
29
 Quantization

Two types of quantization: (a) midtread and (b)

midrise 30
30
 Quantization
• Coding and Decoding
• The process of converting an analog signal
into PCM is called coding, the inverse
operation is called decoding
• Both procedures are accomplished in a CODEC

31
31
Quantization

32
32
Quantization

33
33
Quantization and encoding of a
sampled signal

34
34
 Quantization Error
 When a signal is quantized, we introduce an error -
the coded signal is an approximation of the actual
amplitude value.
 The difference between actual and coded value
(midpoint) is referred to as the quantization error.
 The more zones, the smaller  which results in
smaller errors.
 BUT, the more zones the more bits required to
encode the samples -> higher bit rate

35
35
 Quantization Error (cont.)
• Round-off error
• Overload error
Overload

36
36
 Quantization Noise

Illustration of the quantization process 37

37
 Pulse Code Modulation
• In PCM system,N number of binary digits
are transmitted per sample.Hence the
signaling rate and channel bandwidth of
PCM are very large.
• Also encodind,decoding and quantizing
circuitary of PCM is complex.

38
38
 Delta Modulation
 In Delta Modulation, only one bit is transmitted per
sample
 That bit is a one if the current sample is more positive
than the previous sample, and a zero if it is more
negative
 Since so little information is transmitted, delta
modulation requires higher sampling rates than PCM
for equal quality of reproduction

39
 Delta Modulation
 This scheme sends only the difference between
pulses, if the pulse at time tn+1 is higher in amplitude
value than the pulse at time tn, then a single bit, say a
“1”, is used to indicate the positive value.
 If the pulse is lower in value, resulting in a negative
value, a “0” is used.
 This scheme works well for small changes in signal
values between samples.
 If changes in amplitude are large, this will result in
large errors.

40
40
 Delta Modulation

The process of delta modulation

41
41
 Delta Modulation

Components of Delta Modulation

42
42
 Delta Modulation

DM system. (a) Transmitter. (b) Receiver. 43

43
 Delta Modulation
 Distortions in DM system
1. If the slope of analog signal is much higher than
that of approximated digital signal over long
duration,than this difference is called Slope
overload distortion.
2. The difference between quantized signal and
original signal is called as Granular noise. It is
similar to quantisation noise.

44
44
 Delta Modulation

Two types of quantization errors :

Slope overload distortion and granular noise

45
45
 Delta Modulation
• Distortions in DM system
Granular noise occurs when step size▲ is large relative
to local slope m(t).
There is a further modification in this system,in which
step size is not fixed.
That scheme is known as Adaptive Delta Modulation.

46
46
 Adaptive Delta Modulation
• A better performance can be achieved if
the value of ▲ is not fixed.
• The value of ▲ changes according to the
amplitude of the analog signal.
• It has wide dynamic range due to variable
step size.
• Also better utilisation of bandwidth as
compared to delta modulation.
• Improvement in signal to noise ratio.
47
47
Adaptive Delta Modulation

48
48
 Unit 2
Digital Band-Pass Modulation Techniques
• Digital band-pass modulation techniques
– Amplitude-shift keying
– Phase-shift keying
– Frequency-shift keying
• Receivers
– Coherent detection
• The receiver is synchronized to the transmitter with respect to carrier phases
– Noncoherent detection
• The practical advantage of reduced complexity but at the cost of degraded
performance

50
 Some Preliminaries

• Given a binary source

– The modulation process involves switching ore keying the
amplitude, phase, or frequency of a sinusoidal carrier wave
between a pair of possible values in accordance with symbols 0
and 1.
c(t) A cos(2 f t )
c c c

– All three of them are examples of a band-pass process

1. Binary amplitude shift-keying (BASK)
• The carrier amplitude is keyed between the two possible values used to represent symbols 0
and 1
2. Binary phase-shift keying (BPSK)
• The carrier phase is keyed between the two possible values used to represent symbols 0 and 1.
3. Binary frequency-shift keying (BFSK)
• The carrier frequency is keyed between the two possible values used to represent symbols 0
and 1.

51
– Decreasing the bit duration Tb has the effect of increasing the transmission
bandwidth requirement of a binary modulated wave.

2
Ac 
Tb

2
c ( t)  cos( 2  f t   ) (7 .3 )
c c
Tb

– Differences that distinguish digital modulation from analog modulation.

• The transmission bandwidth requirement of BFSK is greater than that of BASK for
a given binary source.
• However, the same does not hold for BPSK.

52
• Band-Pass Assumption
– The spectrum of a digital modulated wave is centered on the carrier
frequency fc

– Under the assumption fc>>W,

• There will be no spectral overlap in the generation of s(t)
• The spectral content of the modulated wave for positive frequency is essentially
separated from its spectral content for negative frequencies.
s (t )  b (t ) c (t ) (7 .4 )

2
s ( t)  b (t ) cos( 2  f t ) (7 .5 )
c
Tb
– The transmitted signal energy per bit


Tb
E 
2
s (t ) dt
b
0

2 Tb
2
cos 2 ( 2  f ct ) dt

Tb 0
b (t ) (7 .6 )

53
 1
cos ( 2  f c t )  [1  cos( 4  f tc)]
2

2
1 T
1 T
Eb  cos 2 ( 4  f t ) dt
2 2
b (t ) dt 
 
b b

b (t ) c
(7 .7 )
Tb 0
Tb 0

– The band-pass assumption implies that |b(t)|2 is essentially constant

over one complete cycle of the sinusoidal wave cos(4πf t)
Tb
c
cos ( 4  f t ) dt  0
2


2
b (t ) c
0

1 T
Eb 
2


b

b (t ) dt (7 .8 )
Tb 0

– For linear digital modulation schemes governed by Eq.(7.5), the

transmitted signal energy is a scaled version of the energy in the
incoming binary wave responsible for modulating the sinusoidal carrier.

54
 7.2 Binary Amplitude-Shift Keying
– The ON-OFF signaling variety
 E , for binary symbol 1
b(t)
b
(7 .9 )
0, for binary symbol 0

 2E

b
cos( 2  f t c), for symbol 1
s(t) Tb (7 .10 )
0, for symbol 0
– The average transmitted signal energy is ( the two binary symbols must by
equiprobable)

Eb
E av  (7 .11 )
2

55
• Generation and Detection of ASK Signals
– Generation of ASK signal : by using a produce modulator with two inputs
• The ON-OFF signal of Eq. (7.9)
 E , for binary symbol 1
b(t)
b
(7 .9 )
0, for binary symbol 0

• The sinusoidal carrier wave

2
c (t )  cos( 2  f tc )
Tb

– Detection of ASK signal

• The simplest way is to use an envelope detector, exploiting the nonconstant-
envelope property of the BASK signal

56
57
• Spectral Analysis of BASK
– The objective
1) To investigate the effect of varying the carrier frequency fc on the power
spectrum of the BASK signal s(t), assuming that the wave is fixed.
2) Recall that the power spectrum of a signal is defined as 10 times
the logarithm of the squared magnitude spectrum of the signal
3) To investigate the effect of varying the frequency of the square wave on
the spectrum of the BASK signal, assuming that the sinusoidal carrier
wave is fixed.

58
59
1. The spectrum of the BASK signal contains a line
component at f=fc
2. When the square wave is fixed and the carrier
frequency is doubled, the mid-band frequency of the
BASK signal is likewise doubled.
3. When the carrier is fixed and the bit duration is halved,
the width of the main lobe of the sinc function
defining the envelope of the BASK spectrum is
doubled, which, in turn, means that the transmission
bandwidth of the BASK signal is doubled.
4. The transmission bandwidth of BASK, measured in
terms of the width of the main lobe of its spectrum, is
equal to 2/Tb, where Tb is the bit duration.

60
 Phase-Shift Keying
• Binary Phase-Shift Keying (BPSK)
– The special case of double-sideband suppressed-carried (DSB-SC)
modulation
– The pair of signals used to represent symbols 1 and 0,
 2E b
 cos( 2  f t),
c
for symbol 1 correspond ing to i  1
 Tb
s i( t )   ( 7.12 )
2E 2E
 b
cos( 2  f t c  )   b
cos( 2  f t),c for symbol 0 correspond ing to i  2
 Tb Tb

– An antipodal signals
• A pair of sinusoidal wave, which differ only in a relative phase-shift of π
radians.

1. The transmitted energy per bit, Eb is constant, equivalently, the average

transmitted power is constant.
2. Demodulation of BPSK cannot be performed using envelope detection,
rather, we have to look to coherent detection as described next.

61
• Generation and Coherent Detection of BPSK
Signals
1. Generation
– A product modulator consisting of two component
1) Non-return-to-zero level encoder
• The input binary data sequence is encoded in polar form with symbols 1
and 0 represented by the constant-amplitude levels ; √Eb and - √Eb,
2) Product modulator
• Multiplies the level-encoded binary wave by the sinusoidal carrier c(t) of
amplitude √2/Tb to produce the BPSK signal

62
2. Detection
– A receiver that consists of four sections
1) Product modulator; supplied with a locally generated reference signal
that is a replica of the carrier wave c(t)
2) Low-pass filter; designed to remove the double-frequency components of
the product modulator output
3) Sampler ; uniformly samples the output of the low-pass filter, the local
clock governing the operation of the sampler is synchronized with the
clock responsible for bit-timing in the transmitter.
4) Decision-making device ; compares the sampled value of the low-pass
filter’s output to an externally supplied threshold. If the threshold is
exceed, the device decides in favor of symbol 1, otherwise, it decides in
favor of symbol 0.
– What should the bandwidth of the filter be ?
• The bandwidth of the low-pass filter in the coherent BPSK receiver has to
be equal to or greater than the reciprocal of the bit duration Tb for
satisfactory operation of the receiver.

63
64
• Spectral Analysis of BPSK
– The objectives
1. To evaluate the effect of varying the carrier frequency fc on the power
spectrum of the BPSK signal, for a fixed square modulating wave.
2. To evaluate the effect of varying modulation frequency on the power
spectrum of the BPSK signal, for a fixed carrier frequency.

65
66
1. BASK and BPSK signals occupy the same transmission
bandwidth, which defines the width of the main lobe
of the sinc-shaped power spectra.
2. The BASK spectrum includes a carrier component,
whereas this component is absent from the BPSK
spectrum. With this observation we are merely
restating the fact that BASK is an example of
amplitude modulation, whereas BPSK is an example of
double sideband-suppressed carrier modulation
• The present of carrier in the BASK spectrum means that the binary data
stream can be recovered by envelope detection of the BASK signal.
• On the other hand, suppression of the carrier in the BPSK spectrum mandates
the use of coherent detection for recovery of the binary data stream form the
BASK signal

67
• Quadriphase-Shift Keying
– An important goal of digital communication is the
efficient utilization of channel bandwidth
– In QPSK (Quadriphase-shift keying)
• The phase of the sinusoidal carrier takes on one of the four equally
spaced value2s,Esuch as π/4, 3π/4, 5π/4, and 7π/4
 
 cos 2  fc t  ( 2 i  1) , 0  t T
s i( t )     (7 .13 )
T  4
0 , elsewhere

• Each one of the four equally spaced phase values corresponds to a unique
pair of bits called dibit T  2T ( 7.14 )
b

2E   2E  
s(t)  cos ( 2 i  1) cos( 2  f c t )  sin ( 2 i  1) sin( 2  f c t ) (7 .15 )
i
T  4  T  4 
68
1. In reality, the QPSK signal consists of the sum of two BPSK signals
2. One BPSK signal, represented by the first term defined the product of
modulating a binary wave by the sinusoidal carrier
this binary wave has an amplitude equal to ±√E/2
 
2 E / T cos ( 2 i  1) cos( 2  f c t ),
 4 

   E /2 for i  1 , 4
E cos ( 2i  1)   (7 .16 )
 
4   E /2
 for i  2 , 3

3. The second binary wave also has an amplitude equal to ±√E/2

 
 2 E / T sin ( 2 i  1) sin( 2  f c t ),
 4 

    E /2 for i  1 , 2
 E sin ( 2i  1)   ( 7.17 )
 4 for i  3 , 4
  E /2

69
70
• Generation and Coherent Detection of QPSK
Signals
1. Generation
– The incoming binary data stream is first converted into
polar form by a non-return-to-zero level encoder
– The resulting binary wave is next divided by means of
a demultiplexer into two separate binary waves
consisting of the odd- and even- mumbered input bits
of b(t) – these are referred to as the demultiplexed
components of the input binary wave.
– The two BPSK signals are subtracted to produce the
desired QPSK signals

71
72
2. Detection
– The QPSK receiver consists of an In-phase and quadrature with a
common input.
– Each channel is made up of a product modulator, low-pass filter,
sampler, and decision-making device.
– The I- and Q-channles of the receiver, recover the demultiplexed
components a1(t) and a2(t)
– By applying the outputs of these two channels to a multiplexer,
the receiver recovers the original binary sequence

– Each of the two low-pass filters must be assigned a bandwidth

equal to or greater than the reciprocal of the symbol duration T

73
74
• Offset Quadriphase-Shift Keying (OQPSK)
– The extent of amplitude fluctuations exhibited by
QPSK signals may be reduced by using a variant of
quadriphase-shift keying
– The demultiplexed binary wave labeled a2(t) is
delayed by one bit duration with respect to the other
demultiplexed binary wave labled a1(t)
– ±90◦ phase transitions occur twice as frequency but
with a reduced range of amplitude fluctuations.
– Amplitude fluctuations in OQPSK due to filtering have
a smaller amplitude than in QPSK.

75
76
77
• Computer Experiment III : QPSK and OPQSK Spectra
– QPSK Spectra

Carrier Frequency, f c  8 Hz
 1s for part (a) of the figure 
Bit duration, Tb   
 0 s for part (b) of the figure 

– OQPSK Spectra
• For the same parameters used for QPSK

– QPSK occupies a bandwidth equal to one half that of BPSK

78
 Frequency-Shift Keying
• Binary Frequency-Shift Keying (BFSK)
– Each symbols are distinguished from each other by transmitting one of two
sinusoidal waves that differ in frequency by a fixed amount
 2E b
 cos( 2  f t),
1
for symbol 1 correspond ing to i  1
 Tb
s i( t )   (7 .18 )
2E
 b
cos( 2  f t), for symbol 0 correspond ing to i  2
2
 Tb

– Sunde’s BFSK
• When the frequencies f1 and f2 are chosen in such a way that they differ from
each other by an amount equal to the reciprocal of the bit duration Tb

79
• Computer Experiment IV : Sunde’s BFSK
1. Waveform
– Input binary sequence 0011011001 for a bit duration Tb=1s
– The latter part of the figure clearly displays the phase-
continuous property of Sunde’s BFSK
Bit duration, Tb  1s
2. Spectrum
Carrier frequency, f c  8 Hz

1. The spectrum contains two line components at the frequency f=fc±1(2Tb); which equal
7.5Hz and 8.5Hz for fc=8 Hz and Tb=1s
2. The main lobe occupies a band of width equal to (3/Tb)=3Hz, centered on the carrier
frequency fc=8 Hz
3. The largest sidelobe is about 21 dB below the main lobe.

80
81
82
• Continuous-phase Frequency-Shift Keying
– The modulated wave maintains phase continuity at all
transition points, even though at those points in time
the incoming binary data stream switches back and
forth
– Sunde’s BFSK, the overall excursion δf in the
transmitted frequency from symbol 0 to symbol 1, is
equal to the bit rate of the incoming data stream.
– MSK (Minimum Shift Keying)
• The special form of CPFSK
• Uses a different value for the frequency excursion δf , with the result that this
new modulated wave offers superior spectral properties to Sunde’s BFSK.

83
• Minimum-Shift Keying
– Overall frequency excursion δf from binary symbol 1 to symbol 0, is one half
the bit rate f  f  f 1 2

1
 ( 7.19 )
2T b
1
fc  ( f1  f 2 ) ( 7.20 )
2
f
f1  f c  , for symbol 1 ( 7 .21 )
2
f
f2  fc  , for symbol 0 (7 .22 )
2

– Define the MSK signal as the angle-modulated wave

2E b
s (t )  cos[ 2  f t c  ( t )] (7 .23 )
Tb

84
– Sunde’s BFSK has no memory; in other words, knowing which particular
change occurred in the previous bit interval provides no help in the current
bit interval.
f
(t)2  t
 2
t
 , for symbol 1 ( 7.24 )
2T b

 f
 ( t )  2   t
 2
t
  , for symbol 0 (7 .25 )
2T b

85
86
87
88
• Formulation of Minimum-Shift Keying
2Eb 2E b
s (t )  cos(  ( t )) cos( 2  f t )c sin(  ( t )) sin( 2  f t) c ( 7.26 )
Tb Tb

(i)s(t)  E cos(  ( t )) is the in - phase ( I ) component associated with the carrier

I b

2 / T cos( 2  f t ). ( 7.27 )
b c

(ii ) s Q ( t )  E b sin(  ( t )) is the quadrature ( Q ) component associated with the 90  -

phase - shifted carrier. (7 .28 )

 s Q (t ) 
s ( t )  a ( t ) cos( 2  f t )
1

I 1 0
( 7.29 )  ( t )   tan s(t)
 I 

a 2 (t ) 
tan( 2  f t0 ) 
1
s (t )  a (t ) sin( 2  f t )   tan (7 .31 )
(7 .30 )  a ( t)
 1 
Q 2 0

89
1. a2(t)=a1(t)
This scenario arises when two successive binary symbols in the incoming
data stream are the same

 ( t )   tan 1
[tan( 2  f t )]
0

  2 ft ( 7.32 )
0

2. a2(t)=-a1(t)
This second scenario arises when two successive binary symbols in the
incoming data stream are different

 ( t )   tan  1 [  tan( 2  f t )]0

2ft ( 7 .33 )
0

1
f0  (7 .34 )
4T b

90
– Given a non-return-to-zero level encoded binary
wave b(t) of prescribed bit duration Tb and a
sinusoidal carrier wave of frequency fc, we may
formulate the MSK signal by proceeding as follows
1. Use the given binary wave b(t) to construct the binary demultiplexed-
offset waves a1(t) and a2(t)
2. Use Eq. (7.34) to determine the frequency f0
3. Use Eq. (7.29) and (7.30) to determine the in-phase component sI(t) and
quadrature component sQ(t), respectively from which the MSK signal s(t)
follows

91
• Computer Experinment V : MSK Spectrum
– The parameters
Bit duration, T b  1s
Carrier frequency, f c  8 Hz

1. MSK versus QPSK

• The main lobe of MSK occupies a frequency band whose width is 1.5/Tb=1.5Hz
• The transmission bandwidth of MSK is 50 percent larger than that of QPSK
• The sidelobes of MSK are considerably smaller than those of QPSK
2. MSK versus Sunde’s BFSK
• The transmission bandwidth of MSK is one half that of Sunde’s BFSK
• Sunde’s BFSK exhibits two line components at f=fc±1/(2Tb)
• The spectrum of MSK is continuous across the whole frequency band

92
93
– Although the carrier frequency is not high enough
to completely eliminate spectral overlap, the
overlap is relatively small as evidenced by
• The small value of the spectrum at zero frequency
• The small degree of asymmetry about the carrier frequency fc=8Hz

94
 Summary of Three Binary Signaling Schemes

1. BASK, BPSK, and BFSK are the digital counterparts of

amplitude modulation, phase modulation, and
frequency modulation
2. Both BASK and BPSK exhibit discontinuity. It is possible
to configure BFSK in such a way that phase continuity
is maintained across the entire input binary data
stream. The BFSK waveform plotted in part (d) of the
figure is an example of minimum-shift keying

• Table 7.2 presents a summary of the three binary

modulation schemes

95
96
 Non-coherent Digital Modulations Schemes
– Both BASK and BPSK are examples of linear modulation, with increasing
complexity in going from BASK and BPSK.
– BFSK is in general an example of nonlinear modulation

• Noncoherent Detection of BASK Signal

– The system designer would have knowledge of two system parameters
• The carrier frequency fc
• The transmission bandwidth, which is determined by the bit duration Tb.
– The band-pass filter is designed to have a mid-band frequency equal to the
carrier frequency fc and a bandwidth equal to the transmission bandwidth of
the BASK signal.
– The rise time and decay time of the response of the filter to a rectangular
pulse be short compared to the bit duration Tb
1. Band-pass filter produces a pulsed sinusoid for symbol 1, no output for
symbol 0.
2. Envelope detector traces the envelope of the filtered version of the BASK
signal.
3. Decision-making device working in conjunction with the sampler, regenerates
the original binary data stream
97
98
• Noncoherent Detection of BFSK Signals
– The receiver consists of two paths
• Path 1 : uses a band-pass filter of mid-band frequency f1. produce the
output v1
• Path 2 : uses a band-pass filter of mid-band frequency f2. produce the
output v2

– The output of the two paths are applied to a

comparator

99
100
• Differential Phase-Shift Keying
– In the case of phase-shift keying, we cannot have noncoherent
detection in the traditional sense because the term “noncoherent”
means having t to without carrier-phase information
– We employ a “pseudo PSK” technique (differential phase-shift keying)
– DPSK eliminates the need for a coherent reference signal at the
receiver by combination two basic operations at the transmitter
• Differential encoding of the input binary wave
• Phase-shift keying
– The receiver is equipped with a storage capability designed to
measure the relative phase difference between the waveforms
received during two successive bit intervals.
– The phase difference between waveforms received in two succssive bit
intervals will be essentially independent of θ.

101
1. Generation
– The differential encoding process at the transmitter input starts
with an arbitrary first bit, serving merely as reference
• If the incoming binary symbol bk is 1, then the symbol dk is unchanged with respect to the
previous symbol dk-1
• If the incoming binary symbol bk is 0, then the symbol dk is changed with respect to the
previous symbol dk-1

2. Detection
– The phase-modulated pulses pertaining to two successive bits
are identical except for a possible sign reversal
– The incoming pulse is multiplied by the preceding pulse
– The preceding pulse serves the purpose of a locally generated
reference signal
– Applying the sampled output of the low-pass filter to a decision-
making device supplied with a prescribed threshold, detection
of the DPSK signal is accomplished.

102
103
104
 M-ary Digital Modulation Schemes

– we send any one of M possible signals during each

signaling interval of duration T
– The requirement is to conserve bandwidth at the expense
of both increased power and increased system complexity
– When the bandwidth of the channel is less than the
required value, we resort to an M-ary modulation scheme
for maximum bandwidth conservation
• M-ary Phase-Shift Keying
– If we take blocks of m bits to produce a symbol and use an
M-ary PSK scheme with M=2m and symbol duration T=mTb
– The bandwidth required is proportional to 1/(mTb)
– The use of M-ary PSK provides a reduction in transmission
bandwidth by a factor by a factor m=log2M

105
– The discrete coefficients are respectively referred
to as the in-phase and quadrature components of
the M-ary PSK singal
2E  2  i  0,1,..., M  1
s(t)  cos  2  f c t  i , (7 .35 )
i
T  M  0 t T

s ( t )   E cos  2  i    2 
cos( 2  f tc ) 
i M 
    T 
i  0,1,..., M  1
 2  sin( 2  f t )  ,
2
 E sin  i    ( 7 .36)
 c 
  M   T  0tT
1/2

 2   
2

2
 2   
 E cos  i   E sin  i    E, for all i ( 7 .37 )
  M    M  

106
• Signal-Space Diagram
– Pair of orthogonal functions
2
 1( t )  cos( 2  f tc ), 0t T (7 .38 )
T
2
 2 ( t)  sin( 2  f ct ), 0t T (7 .39 )
T

1. M-ary PSK is described in geometric terms by a constellation of M signal points

distributed uniformly on a circle of radius √E
2. Each signal point in the figure corresponds to the signal si(t) of Eq. (7.35) for a
particular value of the index i.
3. The squared length from the origin to each signal point is equal to the signal
energy E.

107
108
• M-ary Quadrature Amplitude Modulation (QAM)
– The mathematical description of the new modulated
signal
2E 2E 0 i  0 ,1,..., M  1
s i( t )  0
a i cos( 2  fct )  b isin( 2  f t),
c
( 7.40 )
T T 0 t T

– The level parameter for in-phase component and

quadrature component are independent of each other
for all I
– M-ary QAM is a hybrid form of M-ary modulation
– M-ary amplitude-shift keying (M-ary ASK)
• If bi=0 for all i, the modulated signal si(t) of Eq. (7.40) reduces to
2E0
s i( t )  a cos(
i
2  f t c) i  0 ,1,..., M  1
T
– M-ary PSK
• If E0=E and the constraint is satisfied

( Ea 2  Eb 2 ) 1 / 2  E, for all i
i i

109
• Signal-Space Diagram
– the signal-space representation of M-ary QAM for
M=16
– Unlike M-ary PSK, the different signal points of M-
ary QAM are characterized by different energy
levels
– Each signal point in the constellation corresponds
to a specific quadbit

110
111
• M-ary Frequency-Shift Keying
– In one form of M-ary FSK, the transmitted signals are defined for some
fixed integer n as

2E   i  0,1,..., M  1
s(t)  cos  ( n  i ) t  , (7 .41 )
T 
i
T 0 t T

– Like M-ary PSK, the envelope of M-ary FSK is constant for all M
T  E for i  j
 0
s i( t ) s j( t )dt  
 0 for i  j
( 7.42 )

– Signal-Space Diagram
• Unlike M-ary PSK and M-ary QAM, M-ary FSK is described by an M-
dimensional signal-space diagram

1 i  0 ,1,..., M  1
 i( t )  s i( t ) (7 .43 )
E 0  t T

112
1. Correlating the signal
2E b
s 1 (t )  cos( 2  f t c) for symbol 1 (7 .45 )
Tb


Tb
s   (t ) s (t ) dt
1 1 1
0

Tb
2
E cos 2 ( 2  f t )dt

 0 Tb
b c
( 7.46 )

s  E (7 .47 )
1 b

under the band-pass assumption,

2E b
s 2( t )   cos( 2  f t )c for symbol 0 (7 .48 )
T
s2   E b b (7 .49 )

2. We may show that the signal

113
114
– As with BPSK, the signal-space diagram consists of
two transmitted signal points
 Eb   0 
s1    (7 .50 ) s2    ( 7 .51 )
 E b
 0 
– Fig. 7.23 and 7.24 differ in one important respect :
dimensionality
2 2
 1( t )  cos( 2  f 1t ) ( 7.52 ) 2 ( t ) cos( 2  f t2 ) (7 .53 )
Tb Tb

1. The separation between the transmitted signal

points for BPSK is √2 times that for BFSK
2. The received signal point lies inside a “cloud”
centered on the transmitted signal point

115
116
 Digital Baseband Transmission
• Why to apply digital transmission?
• Symbols and bits
• Baseband transmission
– Binary error probabilities in baseband transmission
• Pulse shaping
– minimizing ISI and making bandwidth adaptation - cos roll-
off signaling
– maximizing SNR at the instant of sampling - matched
filtering
– optimal terminal filters
• Determination of transmission bandwidth as a function
of pulse shape
– Spectral density of Pulse Amplitude Modulation (PAM)
• Equalization - removing residual ISI - eye diagram
 Why to Apply Digital Transmission?
• Digital communication withstands channel noise, interference
and distortion better than analog system. For instance in PSTN
inter-exchange STP*-links NEXT (Near-End Cross-Talk) produces
several interference. For analog systems interference must be
below 50 dB whereas in digital system 20 dB is enough. With
this respect digital systems can utilize lower quality cabling than
analog systems
• Regenerative repeaters are efficient. Note that cleaning of
analog-signals by repeaters does not work as well
• Digital HW/SW implementation is straightforward
• Circuits can be easily reconfigured and preprogrammed by DSP
techniques (an application: software radio)
• Digital signals can be coded to yield very low error rates
• Digital communication enables efficient exchanging of SNR to
BW-> easy adaptation into different channels
• The cost of digital HW continues to halve every two or three
years

STP: Shielded twisted pair

 Symbols and Bits
D (s y m b o l ra te r  1 / D b a u d )
T ( b it r a t e r  1 / T bi t s /s ) n  lo g 2 M
b b b

1 1 0 1 1 0 0 0 1 1 1 1

s (t )
M 2
n

Generally: s ( t )   a p ( t  k D ) (a PAM* signal)

k
 n : n u m b e r o f b it s

k

For M=2 (binary signalling): s ( t )   a p ( t  k Tb)  M : n u m b e r o f le v e ls

k
k

 D : S y m b o l d u ra tio n
For non-Inter-Symbolic Interference (ISI), p(t) must T : B it d u a ra tio n
1, t  0
b
satisfy:
p(t) 
unipolar,
 0 , t   D ,  2 D ...
2-level pulses
This means that at the instant of decision
s(t)  ak p(tkD) a k
k

*Pulse Amplitude Modulation

 Transmitted power;
bandpass/baseband
Digital
signal BW
Transmissi
Information:
-analog:BW &
dynamic range
Information
source
on
Message
Message
-digital:bit rate estimate Information
sink
Source
Maximization of encoder Source
decoder
information
transferred Channel Channel In baseband systems
Encoder decoder these blocks are missing
Message protection &
channel adaptation;
Interleaving Deinterleaving
convolution, block
coding
Modulator Demodulator
Fights against burst
errors
Received signal
Transmitted
Channel (may contain errors)
M-PSK/FSK/ASK..., signal
depends on channel wireline/wireless
Noise constant/variable
BW & characteristics
Interference linear/nonlinear

• ‘Baseband’ means that no carrier wave modulation is used for transmission

Baseband Digital Transmission Link

t
D
Unipolar PAM

original message bits

received wave y(t)

y ( t )   a k p ( t  t d  k D )  n ( t) decision instances
k

message reconstruction at t  K D  t yields

K d

y (t K )  a k   a k p ( K D  k D )  n ( t )
kK

message ISI Gaussian bandpass noise

 Baseband Unipolar Binary Error
Probability
Assume binary & unipolar x(t)
The sample-and-hold circuit yields:
r .v . Y : y ( t )  a  n ( t )
k k k

Establish H0 and H1 hypothesis:

H : a  0, Y  n
0 k

p (y | H )  p (y)
Y 0 N

and
H : a  1, Y  A  n
1 k

p ( y | H )  p ( y  A)
Y 1 N

pN(y): Noise spectral density

Determining Decision Threshold
H : a  0, Y  n
0 k

p (y | H )  p (y)
Y 0 N

H : a  1, Y  A  n
1 k

p (y | H )  p (yA)
Y 1 N

The comparator implements decision rule:

Choose Ho (a k=0) if Y<V
p  P (Y  V | H )  
V
p (y|H )d y Choose H1 (ak=1) if Y>V
e1 1  Y 1


p eo
 P (Y  V | H )  
0 V
p (y|H
Y 0
)d y

Average error error probability: P e

 P0 P e 0  P P
1 e1
Transmitted „0‟
but detected as „1‟
P  P  1/2  P  1
2 (P  P )
0 1 e e0 e1

Assume Gaussian noise: 2

p (x)
1  x 
e x p  2
 2 
N

 2 
 Determining Error Rate

p e0
  p ( y )dy
V N

2
1 x
   dx
V e x p  

p
e 
 2  2 
0 2

that can be expressed by using the Q-function, defined by

  
2
1 
Q ( k)  exp  d 
k  
2  2 

  V 
2
1 x
 p   dx Q

V e xp
   
2  2  
e0 2

and therefore
V   A V 
pe0  Q   and also P  
V

p N ( y  A )d y  Q 
 

e1 
   
 Baseband Binary Error Rate
in Terms of Pulse Shape and 
setting V=A/2 yields then
    A 
pe  1
2 ( p e0  p e1 ) p p pe  Q  
 2 
e0 e1

for unipolar, rectangular NRZ [0,A] bits

 x  A /2
2 2 2
S R DC

for polar, rectangular NRZ [-A/2,A/2] bits

A/2
 x 
2
A /4
2 2 T /2 2
S 2
2 A  2T  A  A2
a k2       
R DC

0 T  2  2T 2  4
A/ 2 0
and hence
2
 S /( 2 N ), u n ip o la r
2
 A  A
  4N 
R R


 2  R 
S / N , p o la r
R R

  E / N  S / N r  2  N r /( 2 N r )   , u n ip o la r
 
b b 0 R 0b b 0 b 0 b b


N  N r /2
R  2  N r / N r  2  , p o la r
0 b
b 0 b 0 b b

Note that N R
 N 0 B N  N r / 2 (lower limit with sinc-pulses (see later))
0 b
 Pulse Shaping and Band-limited
Transmission

• In digital transmission signaling pulse shape is

chosen to satisfy the following requirements:
– yields maximum SNR at the time instance of decision
(matched filtering)
– accommodates signal to channel bandwidth:
• rapid decrease of pulse energy outside the main lobe in
frequency domain alleviates filter design
• lowers cross-talk in multiplexed systems
 Signaling With Cosine Roll-off Signals
• Maximum transmission rate can be obtained with sinc-pulses
 p ( t )  s i n c  r t   s in c  t / D 

 1  f
 P ( f )  F [ p ( t ) ]    
 r r
• However, they are not time-limited. A more practical choice is the
cosine roll-off signaling:
for raised cos-pulses =r/2

1 f c o s 2  t
  ( f / 2 r)
2
P( f)  r/ 2
cos p ( t)  s in c r t
r 2r 1(4t)
2
 Example
 By using  r/2 and polar signaling, the following waveform
is obtained:

 Note that the zero crossing are spaced by D at

t   0 .5 D ,  1 .5 D ,  2 .5 D , ....

(this could be seen easily also in eye-diagram)

 The zero crossing are easy to detect for clock recovery.
Note that unipolar baseband signaling involves performance
penalty of 3 dB compared to polar signaling:
 Q (  ), u n ip o la r [0 / 1 ]
p  b

e 
 Q ( 2  ), p o la r [  1 ]
b
 Matched Filtering
G (f)
 x ( t ) A p ( t  t ) n


R R 0
y (t )
 X ( f )  A P ( f ) e x p (  j t )
x R (t ) D

R R 0
+ H(f)
  2
2

E   X (f)
R R
df  A
2

R
 P( f ) df
 

A  F [ H ( f) X ( f )]
1

R t  t0  td

 A  H ( f ) P ( f ) e x p  j t df Peak amplitude to be maximized

R d



 

   H(f )
2 2

G ( f )df 
2

n  H ( f ) df Post filter noise

 2  

2


2
 H ( f ) P ( f ) e x p  j t df
A d

  A
2

 Should be maximized

R

2

 H (f ) df
2 
 Matched Filtering SNR and Transfer
Function 2

 Schwartz‟s inequality


 V ( f )W * ( f )d f
 applies when

2

  W ( f ) df



2
 
 V (f ) df V ( f )  KW * ( f )
 

2 
2
 H ( f ) P ( f ) e x p  j t df  SNR at the moment of
A d

   A
2 

 sampling
 
R

2

 H ( f ) df
2  

V(f)H(f ) 
W * ( f )  P ( f ) e x p ( j t ) 
d 
 H ( f )  K P ( f ) e x p ( j t ) d
pulse
impulse response is:  h ( t )  K p( t  t ) energy
d


2
2
2
2 A  P ( f ) df
A
2
Considering right 2A 

  W ( f )
2 R

 
R
df  

hand side yields max SNR    

M AX
Optimum terminal filters
 Px : tra n s m ittin g w a v e fo rm
 Assume 
 H : tra n s m itte r s h a p in g filt e r
– arbitrary TX pulse shape x(t)
T

 H : c h a n n e l t ra n s f e r fu n c tio n
C

– arbitrary channel response h c(t) R : re c e iv e r filt e r

– multilevel PAM transmission
 What kind of filters are required for TX and RX to obtain matched
filtered, non-ISI transmission?

R( f ) Nyqvist shaped
pulse in y(t)
T ( f)
 The following condition must be fulfilled:

Px ( f )HT ( f )H C ( f )R( f ) P(f )exp( jtd )

that means that undistorted transmission is obtained

Avoiding ISI and enabling band-limiting in
radio systems
 Two goals to achieve: band limited transmission & matched filter
reception
TX RX Decision
data
filt. filt. device
T (f ) noise R( f )
T ( f )R( f )C ( f ), raised-cosshaping

N

T ( f )  R *( f ), matchedfiltering
R(f ) T (f ) C( f )
N

 Hence at the transmitter and receiver

alike root-raised cos-filters
must be applied

raised cos-spectra CN(f)

 Determining Transmission Bandwidth
for an Arbitrary Baseband Signaling
Waveform
• Determine the relation between r and B when p(t)=sinc2 at
• First note from time domain that
1, t  0
s in c a t   r a
2

 0 , t   1 / a ,  2 / a...

hence this waveform is suitable for signaling

• There exists a Fourier transform pair
1 f 
s i n c at   1/ a
2

a  a 

B  a
f
• From the spectra we note that T

a a
and hence it must be that
for baseband
 BT  r
PAM Power Spectral Density (PSD)
• PSD for PAM can be determined by using a general expression
Amplitude autocorrelation
1 

 R a( n ) e x p (  j 2  n fD )
2

G (f) 
x
P( f)
D n

• For uncorrelated message bits


2
m ,n0
2 Total power
R(n) a a

 DC power
m ,n 0
a
2

 a

and therefore
 

 R a( n ) e x p (  2  n fD )    m  e x p (  j 2  n fD )
2 2

n a
n  n 

 1 
 n 
on the other hand  e x p (  j 2  n fD )     afnd  r  1/ D
n  
D n  D 


G ( f )  rP(f )  m r  P (nr )  ( f  n r )
2 2
2 2 2

x a a
n
 Example
• For unipolar binary RZ signal:
1 f
P( f)  s in c
2 rb 2 rb
• Assume source bits are equally alike and independent, thus
Tb / 2

  (1 / 2 T )  A d t  A / 4 , m  
2 2 2 2 2

a b a a
0
2 2
A f A  n
 G x ( f )    ( f  n rb ) s i n c
2 2
sin c
1 6r b 2 rb 16 n
2
2

 1 
2
A
r b 
4  2 r  b

G ( f )   r P ( f)
2
2

x a

 m r  P (nr )  ( f  n r )
2
2 2

a
n
 Equalization: Removing Residual ISI
• Consider a tapped delay line equalizer with

p(t)cp(t2ND)
N N

p (t) c p (t )
 N N

p (t) c p ( t  N D)
0 N

• Search for the tap gains cN such that the output equals zero at sample
intervals D except at the decision instant when it should be unity. The
output is (think for instance paths c-N, cN or c0)
N

p ( t )  c p ( t  n D  N D)
eq n
n  N

that is sampled att  k D  N Dyieldingk

N N

p (kDND) 
eq
 cn p ( k D  n D )  cn p D ( k  n ) 
n N n N
 p (t )
Tapped Delay Line:p 0
p 1
p 2 n

Matrix Representation
c  n
c  n 1
c n

p (t )
eq
+
• At the instan t of decision:  1, k  0
p ( t )   c p D ( k  n )   c p
N N


eq k n n 
n N n N kn
 0 , k   1,  2 , ...,  N

 p ... p  c 
 0
• Thatp cleads p cinto (2N+1  ...  p c

0 n
)x(2N+1)0 matri

1 
 n1
x wher    e 
2 n n
0 2 N  N

 ...    
(2N+1
p c ) tap
p c coe
1
 .. f
 n ficien
. p c ts
0
 0can be
 p
 n 1 solv... ed:p  2 n 1 n

N 1
  c   N 1
 0 1

    
 p N ... p  N
 c   1 0
p c  p c  ...  p c  1
n n n 1  n 1 n n
p N 1
... p  N 1
 c  1
 0
    
    
p 2 n c  n  p 2 n 1 c  n  1  ...  p 0 c n  0  p ... p c   0
 2N 0   N  
 Example of Equalization
p
• Read the distorted pulse values into
0

matrix1 .0from
0 .1
fig.
0 .0
(a)
c 0
     p1
0 .1   c    1  p 2
1
 0 .2 1 .0
  0
  p2
 0 .1  0 .2 1 .0   c  0 
  1   p 1

c
1
   0 .0 9 6 
c    0 .9 6 
 0
  
c   0 .2 
   
and the solut ion is 1

Question: what does these

zeros help because they don‟t
exist at the sampling instant? Zero forced values
 Monitoring Transmission Quality
by Eye Diagram

Required minimum bandwidth is

B  r /2
T

Nyqvist‟s sampling theorem:

Given an ideal LPF with the

bandwidth B it is possible to
transmit independent
symbols at the rate:
B r/21/(2T)
T b
 6.6 The Eye Pattern

• Eye Pattern
– Be produced by the synchronized superposition of
successive symbol intervals of the distorted waveform
appearing at the output of the receive-filter prior to
thresholding
– From an experimental perspective, the eye pattern
offers two compelling virtues
• The simplicity of generation
• The provision of a great deal of insightful information about
the characteristics of the data transmission system, hence its
wide use as a visual indicator of how well or poorly a data
transmission system performs the task of transporting a data
sequence across a physical channel.

140
• Timing Features
– Three timing features pertaining to binary data
transmission system,
• Optimum sampling time : The width of the eye opening
defines the time interval over the distorted binary waveform
appearing at the output of the receive-filter
• Zero-crossing jitter : in the receive-filter output, there will
always be irregularities in the zero-crossings, which, give rise
to jitter and therefore non-optimum sampling times
• Timing sensitivity : This sensitivity is determined by the rate
at which the eye pattern is closed as the sampling time is
varied.
Fig. 6.5

141
 Back Next
Fig.6.5

142
• The Peak Distortion for Intersymbol
Interference
– In the absence of channel noise, the eye opening
assumes two extreme values
• An eye opening of unity, which corresponds to zero
intersymbol interference
• An eye opening of zero, which corresponds to a
completely closed eye pattern; this second extreme
case occurs when the effect of intersymbol interference
is severe enough for some upper traces in the eye
pattern to cross with its lower traces.
Fig. 6.6

143
 Back Next
Fig.6.6

144
– Noise margin
• In a noisy environment,
• The extent of eye opening at the optimum sampling time provides a measure of
the operating margin over additive channel noise
( Eye opening )  1  D (6 .34 ) Fig. 6.7
peak

– Eye opening
• Plays an important role in assessing system performance
• Specifies the smallest possible noise margin

• Zero peak distortion , which occurs when the eye opening is unity
• Unity peak distortion, which occurs when the eye pattern is completely closed.
• The idealized signal component of the receive-filter output is defined by the first
term in Eq. (6.10)
• The intersymbol interference is defined by the second term

y 
i
Ea 
i  
k
a p
ik
, i  0 ,  1, 2 ,... ( 6.10 )
k 
k i

145
 Back Next
Fig.6.7

146
 

(Maximum ISI )   pi  k
m  
ki


D peak   pik
k 

  p ( i  k )T b ( 6.35 )
k 
k i

• Eye pattern for M-ary Transmission

– M-ary data transmission system uses M encoded symbols
– The eye pattern for an M-ary data transmission system contains (M-1)
eye openings stacked vertically one on top of the other.
– It is often possible to find asymmetries in the eye pattern of an M-ary
data-transmission system, which are caused by nonlinearities in the
communication channel or other parts of the system.

147
 6.7 Computer Experiment : Eye Diagrams for Binary and
Quanternary Systems
• Fig. 6.8(a) and 6.8(b) show the eye diagrams for a baseband PAM
transmission system using M=2 and M=4.
Fig. 6.8

• Fig. 6.9(a) and 6.9(b) show the eye diagrams for these two
baseband-pulse transmission systems using the same system
parameters as before, but this time under a bandwidth-limited
condition. 1
H ( f )
1(f/ f
2N
0
) Fig. 6.9

1 . N  3 , and f  0 .6 Hz for binary PAM

0

2 . N  3 , and f  0 .3 Hz for 4  PAM

0

B  0 .5 (1  0 .5 )  0 .75 Hz
T

148
 Back Next
Fig.6.8

149
 Back Next
Fig.6.9

150
 Introduction
• What is information theory ?
– Information theory is needed to enable the
communication system to carry information (signals) from
sender to receiver over a communication channel
• it deals with mathematical modelling and analysis of a
communication system
• its major task is to answer to the questions of signal compression
and transfer rate
– Those answers can be found and solved by entropy and
channel capacity
 Entropy
• Entropy is defined in terms of probabilistic
behaviour of a source of information
• In information theory the source output are
discrete random variables that have a certain
fixed finite alphabet with certain probabilities
– Entropy is an average information content for the
given source symbol K1
1
H ( )   p k log 2 ( )
k0 pk
 Entropy (example)
• Entropy of Binary
Memoryless Source
• Binary memoryless
source has symbols
0 and 1 which have
probabilities p0 and
p1 (1-p0)
• Count the entropy
as a function of p0
 Source Coding Theorem
• Source coding means an effective
representation of data generated by
a discrete source
– representation by source encoder
• statistics of the source must be
known (e.g. if coding priorities exist)
 Data Compaction
• Data compaction (a.k.a lossless data
compression) means that we will remove
redundant information from the signal
prior the transmission
– basically this is achieved by assigning short
descriptions to the most frequent outcomes of
the source output and vice versa
• Source-coding schemes that are used in
data compaction are e.g. prefix coding,
huffman coding, lempel-ziv
 Data Compaction example
• Prefix coding has an important feature
that it is always uniquely decodable and
it also satisfies Kraft-McMillan (see
formula 10.22 p. 624) inequality term
• Prefix codes can also be referred to as
instantaneous codes, meaning that the
decoding process is achieved
immediately
 Data Compaction example

• In Huffman coding to
each symbol of a given
alphabet is assigned a
sequence of bits accor -
ding to the symbol
probability
 Data Compaction example

• In Lempel-Ziv coding no probabilities of

the source symbols is needed, which is
actually most often the case
• LZ algorithm parses the source data
stream into segments that are the
shortest subsequences not encountered
previously
Discrete Memoryless Channels
• A discrete memoryless channel is a statistical
model with an input of X and output of Y,
which is a noisy version of X (here both are
random variables)
• In each time slot the channel accepts an input
symbol X selected from a given alphabet
• We can create channel matrix that
corresponds fixed channel inputs and outputs
and we can assume the probabilities of the
symbols
Discrete Memoryless Channels

• Example : Binary symmetric channel

 Mutual Information
• Mutual information uses conditional entropy
of X selected from a known alphabet
– conditional entropy means the uncertainty
remaining about the channel input after the
channel output has been observed
• Mutual information has several properties :
– symmetric channel
– always nonnegative
– relation to the joint entropy of a channel input and
channel output
Mutual Information
 Mutual Information
• Example : Imagine a village in a distance, in
which the inhabitants are nerd-smurfs. One
could divide these smurfs into two distinctive
groups.
– In group A 50% always tells correctly the bitcode if
asked, 30 % will lie if asked, 20 % go beserk
– In group B the same percentages are 30 %, 50 %
and 20%
– calculate mutual information
 Channel Capacity
• Capacity in the channel is defined as a
intrinsic ability of a channel to convey
information
• Using mutual information the channel
capacity of a discrete memoryless
channel is a maximum average mutual
information in any single use of channel
over all possible probability distributions
Discrete Memoryless Channels
• Example : Binary symmetric channel revisited
– capacity of a binary symmetric channel with given
input probabilities
– variability with the error probability
 Channel Coding Theorem

• Channel coding consists of mapping the

incoming data sequence into a channel
input sequence and vice versa via
inverse mapping
– mapping operations performed by encoders

Source Channel Channel Channel Destination

Encoder decoder
Information Capacity Theorem

• A channel with noise and the signal are

received is described as discrete time,
memoryless Gaussian channel (with
power-limitation)
– example : Sphere Packing
 contents
Important classes of error-control coding:

• Linear block codes

• Cyclic codes
• Convolutional codes
• Turbo codes
• Low-density parity-check codes

Communication
Systems Error-Control 168
Coding
 Classification of Error-Control Coding:
Systematic Code Nonsystematic Code

Linear code
Nonlinear code

Random error-correcting code

Burst error-correcting code

Error-detecting code
Error-correcting code

Communication
Systems Error-Control 169
Coding
 Block code
Convolution code

Rule: the message sequence is subdivided into sequential blocks

each k bits long, and each k-bit block is mapped into an n-bit
block, where n>k. the number of redundant bits added by the
channel encoder to each transmitted block is n-k bits.

Code Rate:

k
r 
n r  1
170
Hamming Distance and Hamming Weight
 Hamming weight w(c): defined as the number of nonzero elements in the code vector
c.

10110011 w=5

 Hamming Distance d(c1,c2): defined as the number of locations in which their

respective elements differ between two code words c1 and
c 2.

10110011
d=3
10011010
171
 Minimum Distance and Minimum
Weight
• The minimum weight wmin is defined as the smallest weight of all nonzero code
vectors in the code.

• The minimum distance dmin is defined as the smallest Hamming distance between
any pair of code vectors in the code.

In linear block code d m in  w m in

172
Relation between minimum distance and capability of error-detecting and error-correcting

1 
t  ( d m in  1 )  (1 0 .2 5 )

2 

173
 Figure 10.6
(a) Hamming distance d(ci, cj)  2t  1. (b) Hamming distance
d(ci, cj)  2t. The received vector is denoted by r.

174
 10.3 Linear Block Codes
What is block code?

Message sequence

k bits k bits k bits

encoder

n bits n bits n bits

175
What is Systematic linear block codes?

The message bits are transmitted in unaltered form.

Figure 10.4 Structure of systematic code word

K bits: 1 1 0 1

N bits: 0 0 1 1 0 1
1101 -- systematic code

N bits: 1 0 1 0 0 1-- non-systematic code

176
 k-bit message
m 0 , m 1 , ..., m k 1

n-bit channel code word

c 0 , c 1 , ...c n  1  b 0 , b 1 , ...b n  k  1 , m 0 , m 1 , ..., m k1

parity check bits message bits

The n-k bits are computed from message bits in accordance with a given encoding rule.

177
 Relation between parity bits and message bits

The (n-k) parity bits are linear sums of the k message bits,
as shown by generalized relation

b i  p 0 i m 0  p 1i m 1   p
k  1, i
m
k 1
(1 0 .2 )

Where, the coefficients are defined as follows:

1 if b depends on mj
 
i
p ij
0 o th e r w is e

178
 Matrix Notation
1-by-k message vector m  [m , m , , m k 1 ]
0 1

1-by-(n-k) parity vector b  [b , b , , b n  k 1 ]

0 1

1-by-n code vector c  [c , c , , c n 1 ]

0 1

All of them are row vectors.

It is clear that c can be expressed in terms of m and b as follows

c  [b m ]
179
 bi  p 0 i m 0  p1i m 1   p
k  1, i
m
k 1
(1 0 .2 )

p00 p10 pk 1,0

k-by-(n-k)
 p 00 p0 1 p 0 ,n  k 1 
 
p 10 p 11 p 1, n  k 1
P   
 
 
 p k 1, 0 p k  1 ,1 p k  1 , n  k  1 
b  mP Where pij is 0 or 1.
180
Because

therefore c  [m P m ]  m [ P I ]
k

Ik: k-by-k identity matrix 1 0 0

 
0 1 0
Ik   
 
 
k-by-n generator matrix 0 0 1

G  [P I ] (1 0 .1 2 )
k

Hence , we get generation equation

c  mG (1 0 .1 3 )
181
Why is G called generator matrix?
c  mG

message vector Generator Matrix code vector

m (k bits) G c (n bits)
(channel encoder)

How to get the full set of code words?

The full set of code words, referred to simply as code, is generated in accordance with
c=mG by setting the message vector m range through the set of all 2k binary k-tuples (1-by-k
vectors) 2k.

182
b i  p 0 i m 0  p 1i m 1   p m (1 0 .2 )
k  1, i k 1

183
 Parity-Check Matrix H
Let H denote an (n-k)-by-n matrix, defined as
H  [I PT ]
nk

PT ： (n-k)-by-k matrix, the transpose ofP

since G  [P I ]
k

 PT 
Hence  
HG T
 [ I nk P T]    P T
 P T
 0 Modulo-2
 I  arithmetic
 k 

That is

Systems
HG
Communication
Error-Control
T
 0 184
Coding
 We have known that

c  mG
Postmultiplying both sides by HT, we get

 mGH  0
T T
cH (1 0 .1 6 )

Equation (10.16) is called parity-check equation.

185
 Generator equation
c  mG (1 0 .1 3 )

 0
Parity-check equation T
cH (1 0 .1 6 )
They are basic to the description and operation of a linear block code.

message vector Generator Matrix code vector

m (k bits) G c (n bits)
(channel encoder)

code vector Parity-check Matrix null vector

c (n bits) H 0
(channel decoder)
186
 Repetition Codes

• It is the simplest type of linear block codes.

• (n,1) block code: only 1-bit message, (n-1) parity-check bits, which are the
repetition of the message bit.

187
 Example：
k  1 n  5 I k  [1 ]

c 0 , c 1 , ...c 4  b 0 , b 1 , b 2 , b 3 , m 0

b0  1  m 0

b1  1  m 0 P  [1 1 1 1 ] 1 0 0 0 1
 
0 1 0 0 1
b2  1  m H  [I P ]
T 
nk
0
0 0 1 0 1
 
b3  1  m G  [1 1 1 1 1]
0 0 0 0 1 1

Because: m=0 or 1
Thus, c=mG=00000 or 11111

188
 Error Pattern
transmitted code： c (1×n vector)
received code： r ( 1×n vector)
error pattern： e (1×n error vector)

r  ce e  { e , e , ...e , ...e }

1 2 i n

 1, if a n e rro r h a s o c c u rre d in th e i th lo c a tio n

ei  
0, o th e rw is e
c= [ 0 0 1 1 0 1 0]
If r= [ 0 1 1 1 0 0 0] Then e= [ 0 1 0 0 0 1 0]

Our task now is to decode the code vector c from the received vector r. How to do it?

189
 Definition and properties of
Syndrome
s  rH
Definition: T
(1 0 .1 9 )
HT is a n× (n-k) vector, so, s is a 1×(n-k) vector.

Property 1: it depends only on the error pattern, and not on the transmitted code word.

because:

s  rH  ( c  e) H  cH  eH  eH
T T T T T

190
 Syndrome Decoding
The decoding procedure for a linear block code

① For the received vector r, compute the syndrome s=rHT

② Identify the error pattern emax with the largest probability of occurrence.

③ Compute the code vector c = r + emax as the decoded version of the received vector
r.

191
 Likelihood decoding：

e  { e , e , ...e , ...e }
1 2 i n

s  rH  ( c  e )H  cH  eH  eH
T T T T T

s  eH
T
 e H T
 e H T
 e H T
  e H T
1 1 2 2 i i n n

192
s  eH
T
 e H T
e H T
 e H T
 e H T
1 1 2 2 i i n n

If and only if ei=1, then

s  H sT  H
T or
i i

193
 Example 10.2 Hamming Codes

It is a special (n, k) linear block code that can correct 1-bit error. Speciality :

Block length:
n  2 1
m

Number of message bits:

k  2m  m  1
Number of parity bits:
n k  m
Such as (7, 4) Hamming code.

194
 (7, 4) Hamming Code
 
Generator matrix
1 1 0 1 0 0 0
 
0 1 1 0 1 0 0
G 
1 1 1 0 0 1 0
 
1 0 1 0 0 0 1
 p Ik 

Parity-check matrix  
1 0 0 1 0 1 1
 
H  0 1 0 1 1 1 0
0 0 1 0 1 1 1
 
 Ink
195
pT 
Suppose：received vector is [1100010]
Determine: is it correct?

Because

1 0 0
 
0 1 0
 
0 0 1
 
s  1 1 0 0 0 1 0  1

1 0 

0 0 1
s  0
0 1 1
  So, the received code vector is wrong.
1 1 1
Communication  1 0 1 
Systems Error-Control 
Coding 196
 Cyclic Codes
• Cyclic code is a kind of linear block codes.

• Any cyclic shift of a code word in the code is also a code word.

• Cyclic codes are easy to encode.

• Cyclic codes possess a well-define mathematical structure, which has led to the
development of very efficient decoding schemes for them.

197
 Cyclic Property
Assume that cis a ccode,word of an)(n,k) linear block code.

Code words：

All of them are also code words in the code.

198
 Code Polynomial
Code word:

Code polynomial:

c( X )  c  c X  c X 2
  c X n 1
0 1 2 n1

c (i) ( X )  c   c X
i -1
 c X i
 c X i 1
  c X n -1
(1 0 .3 0 )
n-i n -1
0 1 n - i -1

199
 X c( X )  c X  c X 2
 c X 3
  c X n
0 1 2 n1

X ic ( X )
 X i (c  c X   c X n - i -1
 c X n -i
  c X n-1
)
0 1 n - i -1 n-i n -1

 c X i
 c X i1  c X n -1
 c X n
  c X n  i -1
0 1 n - i -1 n-i n -1

Because in modulo-2 addition,

X c( X )  c  c c Xi c X i 1
 c
i i -1 n-1
X X
n-i n -1 0 1 n - i-1

c (X n
 1)  c X i -1
( Xn 1)
n-i n-1

c ( i ) ( X ) Communication
c   c X
i -1
 c X i
 c X i 1
  c X n -1
(1 0 .3 0 )
Systems n-i
Error-Control n -1
0 1 n - i -1
Coding 200
 Relation between
X i c (X ) a n d c ( i ) ( X )

i (i) remainder
X c( X ) c (X )
 q(X ) 
1 1
n n
X X
Quotient

q ( X ) c c X  c X
i 1
ni n  i1 n 1

X c ( X )  q ( X )( X  1)  c
i n (i)
(X )

( X )  X c ( X ) m o d( X  1)
(i) i n
c (1 0 .3 3 )
Communication
Systems Error-Control 201
Coding
 Generator Polynomial
Let g(X) be a polynomial of degree n-k that is a factor
of Xn+1. It may be expanded as:
nk 1
Generator

nk
g(x)  1  X g i  0 or 1
i
gi X ,
i 1
polynomial

c( X )  a( X )g( X ) (1 0 .3 5 )
202
 Encoding Procedure

Message polynomial m(X )  m  m X   m X k 1

0 1 k 1

c 0 , c 1 , ...c n  1  b 0 , b 1 , ...b nk1

, m 0 , m 1 , ..., m k1

n-k parity check bits k message bits

b( X )  b  b X   b X n  k 1
0 1 n  k 1
Therefore, code polynomial

nk
Systems
X )  b( X )  X
c Error-Control
(Communication 203
m( X ) ( 1 0 .3 8 )
Coding
 How to determine b(X)?
Because c ( X )  a ( X ) g (X ) (1 0 .3 5 )
nk
c(X )  b(X )  X m (X ) ( 1 0 .3 8 )

Thus, nk
a(X)g(X)  b(X)  X m(X)

nk
X m(X ) b( X )
 a(X )  (1 0 .3 9 )
g(X ) g(X )
It states that the polynomial b(X) is the remainder left over after dividing xn-k m(X) by g(X).

204
 Steps for encoding an (n,k)
systematic cyclic code

1. Multiply the message polynomial m(X) by Xn-k;

2. Divide Xn-k m(X) by the generator polynomial g(X), obtaining the remainder b(x);
3. Add b(x) to Xn-k m(X) , obtaining the code polynomial c(X).

205
How to select generator polynomial?

Rule: Any factor of Xn+1 can be used as a generator polynomial, the degree of the factor
determines the number of parity bits.

It is difficult to select a polynomial factor to construct a GOOD cyclic code.

206
 Example: (7, k) cyclic code
3
 1  (1  X ) (1  X  X ) (1  X  X )
7 2 3
X
(7,6) cyclic code

(7,4) cyclic code

(7,4) cyclic code

(7,3) cyclic code

(7,3) cyclic code

(7,1) cyclic code

207
 Generator and Parity-Check Matrices
Generator matrix polynomial

 g(X ) 
  coefficients
Xg(X )
G(X )    G H
 
 k 1 
X g(X )

G  [P I ]
k

Note: G produced by this way may not be systematic .

208
 Encoder for Cyclic Codes

How to design the cyclic encoder?

These encoding procedure can be implemented by means of encoder consisting of

a linear feedback shift register with (n-k) stages.

209
 Fig.10.8 Encoder for an (n, k) cyclic code

D D D D D

n  k  1


n  k
g ( x)  1    1 o r 0
i
g i
X X , g i
i  1

m  [m , m , ,m ]
0 1 k1

...c n  1  b 0 , b 1 , ...b n  k  1 , m 0 , m 1 , ..., m

c 0 , c 1 ,Communication k1
Systems Error-Control
Coding 210
 Operation of the encoder
1. The gate is switched on. The k message bits are shifted into the channel, and
enter the shift registers.
2. The gate is switched off, thereby breaking the feedback connections.
3. The contents of the shift register are read out into the channel.

Communication
Systems Error-Control 211
Coding
 Calculation of the Syndrome

T
s r H

• If the syndrome is zero, there are no transmission errors in the received word.
• If the syndrome is nonzero, the received word contains transmission errors that
require correction.

In the case of a cyclic code in systematic form, the syndrome can be calculated
easily.
Communication
Systems Error-Control 212
Coding
 Syndrome polynomial
Suppose： Code word

Received code

Then, the received code polynomial

Quotient q(X)
r(X )
Remainder s(X)
g(X )

r( X )  q( X ) g( X )  s( X ) (1 0 .4 7 )

213
 Syndrome Calculator for (n,k) Cyclic Code

D D D

• Syndrome calculator is identical to the encoder.

• As soon as all the received bits have been shifted into the shift register, its contents are
the syndrome S.

214
 In this example, we have two primitive polynomials:

1 X  X
2 3
1 X  X
3

Here, we take
g ( X )  (1  X  X )
3

Hence, parity-check polynomial

1
7
X
h( X )   (1  X )(1  X  X )
2 3

g ( x)

1 X  X 
2 4
215 X
 Suppose: message sequence 1001

? Determine: the whole code word xxx1001

Message polynomial

nk
m(X )  X  X
3 6
m(X )  1 X
3
X

 X X  X
3 6 2
X
 X  X 
3

1 X  X 1 X  X
3 3

Quotient remainder
a(X )  X  X
3
b( X )  X  X
2

Code polynomial

nk
c( X )  b( X )  X m(X )  X  X  X  X
2 3 6

Code word 0111001 216

 ?
Suppose: g(X)=1+X+X3 is given
Determine: G and H

g ( X ) 1  X  X
3
 
g(X )
 
Xg (X )  X  X  X
Xg(X ) 2 4
G(X )   
  3 5
g(X )  X  X  X
2 2
 k1  X
X g ( X )
X g(X )  X  X  X
3 3 4 6

1  X  X 
3
1 1 0 1 0 0 0
  0
X  X  X 0
2 4
1 1 0 1 0
G(X)   G'   
  0 0
 X  X 0 1 1 0 1
2 3 5
X  
  0 0 0 1 1 0 1
X  X  X 
3 4 6

Communication
Systems Error-Control 217
Coding
 1 1 0 1 0 0 0 1 1 0 1 0 0 0
   
0 1 1 0 1 0 0 0 1 1 0 1 0 0
'
G   G   
0 0 1 1 0 1 0 1 1 1 0 0 1 0
   
0 0 0 1 1 0 1 1 0 1 0 0 0 1

Non-systematic form Systematic form

If m=(1001) C=0111001

G  [P Ik ] ( 1 0 .1 2 ) H [I P ]
T
( 1 0 .1 4 )
n k

1 0 0 1 0 1 1
H  0 1 0 1 1 1 0

 
0 0 1 0 1 1 1 
218
Suppose: g(X)=1+X+X3 is given
Determine: encoder and syndrome calculator

n  k 1


nk
g ( x )  1  X
i
gi X
i 1

g(X)=1+X+X3
g1  1, g2  0

D D D

219
 D D D

1001

shift input Register contents

0 0 0 initial state
1 1 110 1001

2 0 011
0111001
3 0 111
4 1 0 1 1 parity bits
229
 Suppose: received code is 0110001
Determine: is it correct?

1 0 0
 
0 1 0
 
0 0 1
 
s  rH  0 1 1 0 0 0 1   1 1 0
T
1 0  1
0 1 1
 
1 1 1
1 1 
 0

s  0 So, the received code vector is wrong.

221
Syndrome calculator

0110001
D D D
Fig 10.11

Shift Input Register contents

0 0 0 (initial state)
1 1 100
2 0 010
3 0 001 s  0
4 0 110
5 1 111 So, the received code vector is wrong.

6 1 001
7Systems 0Communication 1 1 0 syndrome
Error-Control 222
Coding
Convolutional codes
Introduction
• Convolutional codes map information to code bits sequentially by
convolving a sequence of information bits with “generator”
sequences

• A convolutional encoder encodes K information bits to N>K code bits

at one time step

• Convolutional codes can be regarded as block codes for which the

encoder has a certain structure such that we can express the
encoding operation as convolution

- 224 -
Properties of convolutional codes
Consider a convolutional encoder. Input to the encoder is a
information bit sequence u(partitioned into blocks of length K):
u  (u 0 , u1 ,)
(1) (2) ( K)
ui  (ui ,ui ,ui ),
The encoder output the code bit sequence x(partitioned into blocks
of length N)
x  ( x 0 , x1 ,)
(1) (2) (N )
xi  (xi , xi ,xi ),

K
R
N
- 225 -
Example: Consider a rate ½ convolutional code with K=1 and N=2 defined by
the circuit:
xi
ui

(1)
The sequences (x 0 , x1(1) ,), (x0(2) , x1(2) ,) are generated as follows:

xi(1) u i and x(2)  u u

i i i1

Multiplexing between x(1)and x(2)gives the code bit sequence

i i

x  ((x0(1) x0(2) ),(x1(1) x1(2) ),)  (x 0 , x1,)

- 226 -
 • The convolutional code is linear

• The encoding mapping is bijective

• Code bits generated at time step i are affected by information bits up

to M time steps i – 1, i – 2, …, i – M back in time. M is the maximal
delay of information bits in the encoder

• Code memory is the (minimal) number of registers to construct an

encoding circuit for the code.

• Constraint length is the overall number of information bits affecting

code bits generated at time step i: =code memory + K=MK + K=(M +
1)K

• A convolutional code is systematic if the N code bits generated at

time step i contain the K information bits

- 227 -
 Example: The rate ½ code defined by the circuit
xi
ui

has delay M=1, memory 1, constraint length 2, and it is systematic

Example: the rate 2/3 code defined by the circuit

x i(1)
u i(1)
x i(2)
x i(3)
u i(2)

has delay M=1, memory 2, constraint length 4, and not systematic

- 228 -
 Tree u u1 u
0 2

00
00
A
11
00 A

01
11
B
10
1
A
00
0 01
A
11
11

01
10
B
10

- 229 -
Trellis
The tree graph can be contracted to a direct graph called trellis
of the convolutional code having at most S nodes at distance
i=0,1,… to the root

The contents of the (at most) MK encoder registers are assigned

the variables s ( j) GF(2), j  0,1,,M  K 1
i

(MK 1)
The vector s i  (s i ,si ,si
(0) (1)
)
combibing all register contents at time step i is called state of
the encoder at time step i.

The code bit block x iis clearly a function of s iand ui, only
- 230 -
 Example:

The encoder of the rate ½ convolutional code has S  21  2

different states. The state is given by
s
The code bit block xiat time step i is computed from siand uiby
x(1)
 uand x(2)  u  s
i i i i i

00 00 00
A A A A
11 11 11 ui  0
ui 1
01 01
B B B
10 10
- 231 -
 Example: Constructing a trellis section
xi
ui

x i(1)  u i and x (i 2 )  u i  s i
Two equations are required:

(1) How does s i depend on u im and possibly s i  m , m0

s i  u i1
(2) How does x i depend on s i and u i
x i(1)  u i and x (i 2 )  u i  s i
The branches are labeled with u i | xi called state transition
leading from a state si to a new state s i 1

- 232 -
 Trellis section:
0|00
0 0

0|01 1|11

1 1
1|10

s0 s1 s2 s L2 s L1
0|00 0|00 0|00
0 0 0 0 0

0|01 1|11 0|01 1|11 0|01 1|11

1 1 1 1 1
1|10 1|10 1|10
- 233 -
 s0 s1 s2 s L2 s L1
0|00 0|00 0|00
0 0 0 0 0

1|11 0|01 1|11 0|01 1|11

1 1 1 1
1|10 1|10

- 234 -
 State diagram
Example: Trellis of the rate ½ convolutional code
x i(1)  u i and x i(2)  u i  s i
si s i 1
00
0 0
ui  0
01 11 ui  1
1 1
10

State diagram:
00
0
ui  0
11 01 ui  1
1
10
- 235 -
 Description with submatrices
Definition: A convolutional code is a set C of code bit sequences

x  (x0 , x1,, xi ,),

x i  (xi , xi ,xi
(1) (2) (N)
), x ( j)
i GF(2)
There exist many encoders mapping information bit sequences
partitioned into lenth N blocks

u  (u 0 , u1 ,)
( j)
ui  (ui ,ui ui GF(2)

(partitioned into length K<N blocks) to code bit sequences x

for the same code
- 236 -
 Example: the following two encoding curcuits generate the
same set of code word sequences

xi(1) xi(1)
ui(1) ui(2)
xi(2) xi(2)
xi(3) xi(3)
ui(2) ui(1)

- 237 -
 Generator matrix

x  u G,
where

G 0 G1 G2  G M 
 
 G0 G1 G 2  GM 
G  
G0 G1 G 2  GM 

   
M
xi  u im G m ,i
m0
The generated convolutional code has rate R=K/N, memory
K*M, and constraint length K*(M+1)
- 238 -
 Example:
The rate ½ code is given by
xi(1) u i and xi(2)  ui  si
G0 governs how ui affects xi  (x x ) : G0  1 1
(1) (2)
i i

G1 governs how ui1affects xi : G1  0 1

((x(1) x(2) ),(x(1) x(2) ),(x(1) x(2) ))  (u ,u ,u )G,

0 0 1 1 2 2 0 1 2

where
11 01 
 
G 11 01
 
 11 
- 239 -
 Description with polynomials
 g1(1) (D) g1(2) (D)  g1(N) (D)
 
 g 2 (D) g 2 (D)  g 2 (D)
(1) (2) (N)
G(D)   
    
 g (1) (D) g(2) (D)  g (N) (D) 
 K K K 
gi(j) (D)gi,0
(j)
gi,1( j) D1 gi,2 D gi,M
( j) 2
D ,gi,m GF(2)
( j) M (j)

u(D)  (u (1) (D),u(2) (D),u (K ) (D))where

u ( j) (D)  u0( j) u1( j) Dui( j) Di , j 1,2,,K,
x(D)  (x(1) (D), x(2) (D),, x( N ) (D)) where
x ( j) (D)  x 0( j)  x1( j) D  xi( j) Di , j  1,2,,N

x(D)  u(D)G(D)
( j)
gi,m  G m (i, j),i 1,, K, j 1,,N, m  0,,M,

M  maxdeg(g i( j) (D)),i 1,,K, j  1,,N

i, j

- 240 -
 Example:
The rate ½ code is given by
x u i and i  u i  s i
(1)
i
x (2)

g1(1) (D)and g1(2) (D) G(D)(g1(1) (D)g1(2)(D))

deg((g i( j) (D))) 1
From M=1 follows that

The polynomial g (1) (D) governs how ulm , m 0,1, affects

1

1 (D) 1 0 D 1
xl(1) : g (1)
The polynomial
1 g (D) governs how ulm , m 0,1, affects
(2)

xl(2) : g 1(2) (D)  11D  D 1

u  (1,1,0,)  u(D)  1 D
x  u G,yielding x  (11,10,01,00,)
x(D)  u(D)G(D)yielding x(D)  1 D 1 D  2

- 241 -
 Decoding of convolutional codes
The Viterbi algorithm

x1j
uj
s1j s2 j
x2 j

u  (1,1,1,1,1,1,1)
(s10 , s20 ) (1,1)
x  (11,11,11,11,11,11,11)

- 242 -
 +1/+1+1
+1+1 +1+1
-1/-1-1

+1/+1-1
+1-1 +1-1
-1/-1+1

+1/+1-1
-1+1 -1/-1+1 -1+1

+1/+1+1
-1/-1-1
-1-1 -1-1
s1 j ,s2 j x1 j , x2 j s1j1 ,s2 j1
- 243 -
 Hard decisions
y (+1+1,-1+1,+1+1,+1+1,+1+1,+1+1,+1+1)
j  x1(m)   2 j y
(m) (m)
j y 1j x 2j

+1+1 -1+1 +1+1 +1+1 +1+1 +1+1 +1+1

+1+1 +2 +2 0 +2 +2 +4 +6 +8 +10 +12

0 -2 0 -2
-2 +2 0 0 +2 +4 +6 +8
0
-2
0
-4 +2 +4 +2 +4
+2 +6
0
+2
0 -2 +2 0 +2 +4 +6

j=0 j=1 j=2 j=3 j=4 j=5 j=6 j=7

- 244 -
 +1+1 -1+1 +1+1 +1+1 +1+1 +1+1 +1+1

+1+1 +2 +2 0 +2 +2 +4 +6 +8 +10 +12

0 -2 0 -2
-2 +2 0 0 +2 +4 +6 +8
0
-2
-4 0 +2 +4 +2 +4
+2 +6
0
+2
0 -2 +2 0 +2 +4 +6

j=0 j=1 j=2 j=3 j=4 j=5 j=6 j=7

uˆj +1 +1 +1 +1 +1 +1 +1 - No error

- 245 -
 Hard decisions
y (+1+1,-1+1,+1-1,+1+1,+1+1,+1+1,+1+1)
j  x 1( mj )  y1 j  x 2( mj )  y
(m)
2j

+1+1 -1+1 +1-1 +1+1 +1+1 +1+1 +1+1

+1+1 +2 +2 0 +2 0 +2 +2 +4 +6 +8 +2 +10
0 -2 0 0 -2
-2 +2 +2 +2 0 +4 +2 +4 +6
-2 0
-2
+2 +4 0
-4 +2 +4 +6
+2 +4
-2 0

0 +2
0 0 0 -2 +2 +4 +2 +4

j=0 j=1 j=2 j=3 j=4 j=5 j=6 j=7

- 246 -
 +1+1 -1+1 +1-1 +1+1 +1+1 +1+1 +1+1

+1+1 +2 +2 0 +2 0 +2 +2 +4 +6 +2 +8 +2 +10
0 -2 0 0 -2
-2 +2 +2 +2 0 +4 +6
-2 0
-2
+2 +4 0
-4 +2
+2 +4
-2 0

0 +2
0 0 0 -2 +2 +4

j=0 j=1 j=2 j=3 j=4 j=5 j=6 j=7

uˆj +1 +1 +1 +1 +1 +1 +1 - No error

- 247 -
 Hard decisions
y (+1+1,-1-1,-1+1,+1+1,+1+1,+1+1,+1+1)
j  x1(m)   2 j y 2j
(m) (m)
j y 1j x

+1+1 -1-1 -1+1 +1+1 +1+1 +1+1 +1+1

+1+1 +2 +2 -2 0 0 0 +2 +2 +8 +10 +12

0 -2 +2 0 -2
-2 +4 -2 0 0 +2 +8 +6 +8
+2 0
0
-2 +2 0
-2 +8 +6 +8
0 +10
+2 0
0 +2
-2 0 +6 -2 +4 +8 +6

j=0 j=1 j=2 j=3 j=4 j=5 j=6 j=7

- 248 -
 +1+1 -1-1 -1+1 +1+1 +1+1 +1+1 +1+1

+1+1 +2 +2 -2 0 0 0 +2 +2 +2 +8 +2 +10 +12

0 -2 +2 0 -2 -2
-2 +4 -2 0 0 +2 0 +6 +8
+2 0
0
-2 +2 0
-2 +8 +8
0 +10
+2 0
0 +2 +2
-2 0 +6 -2 +4 +8 -2 +6

j=0 j=1 j=2 j=3 j=4 j=5 j=6 j=7

uˆj +1 -1 -1 +1 +1 +1 +1 - 2 decoding errors

- 249 -
Soft decisions
  2, x(m)
ij  y ij, GOOD channel

 1/ 2, x (m)
 yij , BAD channel
lij   ij

 1/2, x (m)
ij  yij , BAD channel
  2,xij(m)  yij , GOOD channel

(m)  x l y1j  x2 j l2 j y2 j
j 1 j 1j

CSI values ((G,B),(B,B),(G,G),(G,B),(B,B),(G,G),(G,G))

y (+1+1,-1-1,-1+1,+1+1,+1+1,+1+1,+1+1)
- 250 -
 +1+1 -1-1 -1+1 +1+1 +1+1 +1+1 +1+1
GB BB GG GB BB GG GG
+2+0.5 -0.5-0.5 -2+2 +2+0.5 +0.5 +0.5 +2+2 +2+2

+1+1 +2.5 +2.5 +1 +3.5 +4 +7.5 +2.5 +10 +1 +11 +4 +14 +4 +18
0 -2.5 -1 -4 -2.5 -1 -4 -4
-2.5 +1.5 0 -0.5 0 +5 0 +9 0 +7 0 +10
0 0 0 0 0
0
0 0 0 0 0
-2.5 +1.5 +4 +5 +9
0 +13
0 0 0 0 0
+2.5
+4 +1 +4 +4
-2.5 -4 +1.5 -2.5 -0.5 -1 +5 -4 +9 -4 +7

j=0 j=1 j=2 j=3 j=4 j=5 j=6 j=7

- 251 -
 +1+1 -1-1 -1+1 +1+1 +1+1 +1+1 +1+1
GB BB GG GB BB GG GG
+2 +0.5 -0.5-0.5 -2 +2 +2 +0.5 +0.5 +0.5 +2 +2 +2 +2

+1+1 +2.5 +2.5 +1 +3.5 +4 +7.5 +2.5 +10 +1 +11 +4 +14 +4 +18
0 -2.5 -1 -4 -2.5

-2.5 +1.5 0 -0.5 0 +10

0 0
0
0 0
-2.5 +1.5
0 +13
0 0
+2.5
+4
-2.5 -4 +1.5 -2.5 +7

j=0 j=1 j=2 j=3 j=4 j=5 j=6 j=7

uˆj +1 +1 +1 +1 +1 +1 +1 - No error

- 252 -
Spread Spectrum
Introduction to Spread Spectrum

• Problems such as capacity limits, propagation

effects, synchronization occur with wireless systems
• Spread spectrum modulation spreads out the
modulated signal bandwidth so it is much greater
than the message bandwidth
• Independent code spreads signal at transmitter and
despreads signal at receiver
 Multiplexing
channels ki
• Multiplexing in 4 dimensions
– space (si) k1 k2 k3 k4 k5 k6

– time (t) c
– frequency (f) t c
– code (c) t
s1
f
• Goal: multiple use s2
f
of a shared medium
c
t
• Important: guard spaces needed!
s3
f
 Frequency multiplex
• Separation of spectrum into smaller frequency bands
• Channel gets band of the spectrum for the whole time
• Advantages:
k3 k4 k5 k6
– no dynamic coordination needed
– works also for analog signals c
f
• Disadvantages:
– waste of bandwidth
if traffic distributed unevenly
– inflexible
– guard spaces

t
 Time multiplex
• Channel gets the whole spectrum for a certain
amount of time
• Advantages:
– only one carrier in the
medium at any time
– throughput high even k1 k2 k3 k4 k5 k6
c
for many users
f
• Disadvantages:
– precise
synchronization
necessary
t
 Time and frequency multiplex
• A channel gets a certain frequency band for a certain
amount of time (e.g. GSM)
• Advantages:
– better protection against tapping
– protection against frequency
selective interference
– higher data rates compared to k1 k2 k3 k4 k5 k6
code multiplex c
• Precise coordination f

required

t
 Code multiplex
k1 k2 k3 k4 k5 k6

• Each channel has unique code

c
• All channels use same spectrum at same time
• Advantages:
– bandwidth efficient
– no coordination and synchronization
– good protection against interference
• Disadvantages: f
– lower user data rates
– more complex signal regeneration
• Implemented using spread spectrum technology

t
 Spread Spectrum Technology
• Problem of radio transmission: frequency dependent
fading can wipe out narrow band signals for duration of
the interference
• Solution: spread the narrow band signal into a broad
band signal using a special code

interference
spread power signal
power signal spread
interference
detection at
receiver
f f
 Spread Spectrum Technology
• Side effects:
– coexistence of several signals without dynamic coordination
– tap-proof
• Alternatives: Direct Sequence (DS/SS), Frequency
Hopping (FH/SS)
• Spread spectrum increases BW of message signal by a
factor N, Processing Gain

B
 B ss 
P r o c e s s in g G a i n N  ss
 1 0 lo g 1 0  
B  B 
Effects of spreading and interference

user signal
broadband interference
narrowband interference
P P

i) ii)
f f
P sender P P

iii) iv) v)
f f f
receiver
Spreading and frequency selective
fading
channel
quality

2
narrowband
1 5 6
3 channels
4

Narrowband frequency
guard space
signal
channel
quality
2
2
2
2
2
1 spread spectrum
channels
spread frequency
spectrum
 DSSS (Direct Sequence Spread Spectrum) I

• XOR the signal with pseudonoise (PN) sequence (chipping sequence)

• Advantages
– reduces frequency selective
fading Tb
– in cellular networks user data
• base stations can use the
same frequency range
0 1 XOR
• several base stations can Tc
detect and recover the signal chipping
• But, needs precise power control sequence
01101010110101
=
resulting
signal
01101011001010
 DSSS (Direct Sequence Spread Spectrum) II

transmitter
Spread spectrum
Signal y(t)=m(t)c(t) transmit
user data signal
X modulator
m(t)
chipping radio
sequence, c(t) carrier

receiver correlator
sampled
received products data
sums
signal demodulator X integrator decision
radio
carrier
Chipping sequence,
c(t)
 DS/SS Comments III
• Pseudonoise(PN) sequence chosen so that its
autocorrelation is very narrow => PSD is very
wide
– Concentrated around  < Tc
– Cross-correlation between two user’s codes is very
small
 DS/SS Comments IV
• Secure and Jamming Resistant
– Both receiver and transmitter must know c(t)
– Since PSD is low, hard to tell if signal present
– Since wide response, tough to jam everything
• Multiple access
– If ci(t) is orthogonal to cj(t), then users do not interfere
• Near/Far problem
– Users must be received with the same power
 FH/SS (Frequency Hopping Spread
Spectrum) I
• Discrete changes of carrier frequency
– sequence of frequency changes determined via PN sequence
• Two versions
– Fast Hopping: several frequencies per user bit (FFH)
– Slow Hopping: several user bits per frequency (SFH)
• Advantages
– frequency selective fading and interference limited to short period
– uses only small portion of spectrum at any time
• Disadvantages
– not as robust as DS/SS
– simpler to detect
FHSS (Frequency Hopping Spread
Spectrum) II
Tb

user data

0 1 0 1 1 t
f
Td
f3
slow
f2 hopping
(3 bits/hop)
f1

Td t
f

f3 fast
f2 hopping
(3 hops/bit)
f1

t
Tb: bit period Td: dwell time
Code Division Multiple Access (CDMA)
• Multiplexing Technique used with spread spectrum
• Start with data signal rate D
– Called bit data rate
• Break each bit into k chips according to fixed pattern specific
to each user
– User’s code
• New channel has chip data rate kD chips per second
• E.g. k=6, three users (A,B,C) communicating with base
receiver R
• Code for A = <1,-1,-1,1,-1,1>
• Code for B = <1,1,-1,-1,1,1>
• Code for C = <1,1,-1,1,1,-1>
CDMA Example
 CDMA Explanation
• Consider A communicating with base
• Base knows A’s code
• Assume communication already synchronized
• A wants to send a 1
– Send chip pattern <1,-1,-1,1,-1,1>
• A’s code
• A wants to send 0
– Send chip[ pattern <-1,1,1,-1,1,-1>
• Complement of A’s code
• Decoder ignores other sources when using A’s code to decode
– Orthogonal codes
 CDMA for DSSS
• n users each using different orthogonal PN
sequence
• Modulate each users data stream
– Using BPSK
• Multiply by spreading code of user
CDMA in a DSSS Environment
Seven Channel CDMA Encoding and
Decoding
FHSS (Frequency Hopping Spread Spectrum) III

transmitter narrowband Spread transmit

signal signal
user data
modulator modulator

frequency hopping
synthesizer sequence
receiver

received data
signal demodulator demodulator

hopping frequency
sequence synthesizer
 Applications of Spread Spectrum
• Cell phones
– IS-95 (DS/SS)
– GSM
• Global Positioning System (GPS)
• Wireless LANs
– 802.11b
 Performance of DS/SS Systems
• Pseudonoise (PN) codes
– Spread signal at the transmitter
– Despread signal at the receiver
• Ideal PN sequences should be
– Orthogonal (no interference)
– Random (security)
– Autocorrelation similar to white noise (high at =0
and low for  not equal 0)
 PN Sequence Generation
• Codes are periodic and generated by a shift register and XOR
• Maximum-length (ML) shift register sequences, m-stage shift
register, length: n = 2m – 1 bits

R()

 
-1/n Tc nTc
-nTc

Output
+
 Generating PN Sequences

Output m Stages connected

+ to modulo-2 adder
• Take m=2 =>L=3 2 1,2
• cn=[1,1,0,1,1,0, . . .],
usually written as bipolar 3 1,3
cn=[1,1,-1,1,1,-1, . . .]
4 1,4
5 1,4
1 L

R c m   c n
cnm 6 1,6
L n 1

1 m 0 8 1,5,6,7
 
  1/ L 1  m  L 1
 Problems with m-sequences
• Cross-correlations with other m-sequences
generated by different input sequences can be
quite high
• Easy to guess connection setup in 2m samples
so not too secure
• In practice, Gold codes or Kasami sequences
which combine the output of m-sequences
are used.
 Detecting DS/SS PSK Signals
transmitter
Spread spectrum
Signal y(t)=m(t)c(t) transmit
Bipolar, NRZ signal
m(t) X X

PN
sequence, c(t) sqrt(2)cos (ct + )

receiver
received z(t) w(t) data
signal
X X LPF integrator decision
x(t)

sqrt(2)cos (ct + ) c(t)

 Optimum Detection of DS/SS PSK
• Recall, bipolar signaling (PSK) and white noise give
the optimum error probability
 2 Eb 
Pb  Q  

  

• Not effected by spreading

– Wideband noise not affected by spreading
– Narrowband noise reduced by spreading
 Signal Spectra

B
 B ss  T
P r o c e s s in g G a in N  ss
 1 0 lo g 10  b
 
B  B  Tc

• Effective noise power is channel noise power

plus jamming (NB) signal power divided by N

Tb

Tc
 Multiple Access Performance
• Assume K users in the same frequency band,
• Interested in user 1, other users interfere

4 6

1
3 2
 Signal Model
• Interested in signal 1, but we also get signals
from other K-1 users:
x k t   2 m k  t   k  c k  t   k  c o s  c  t   k    k 

 2 m k  t   k  c k  t   k  c o s  c t   k   k   k   c k
• At receiver,

K
x  t   x1  t    x k t 
k2
 Interfering Signal

• After mixing and despreading (assume 1=0)

z k t   2 m k t   k  c k t   k  c1 t  c o s  c t   k  c o s  c t   1 
• After LPF
w k t   m k t   k  c k t   k  c1 t  c o s  k  1 

• After the integrator-sampler

Tb
Ik  c o s  k   1   m
k  t   k  c k  t   k  c1  t  d t
0
 At Receiver
• m(t) =+/-1 (PSK), bit duration Tb
• Interfering signal may change amplitude at k
 k Tb 
I k  c o s  k   1   b  1  ck  t   k  c1  t  d t b 0  c k  t   k  c1  t  d t 
0
 k 

• At User 1: I1 
Tb
0 m
1  t  c 1  t  c1  t  d t
• Ideally, spreading codes are Orthogonal:
Tb Tb
0 1  t  c 1  t d t  A 0 t   k  c 1 t  d t  0
c c
k
 Multiple Access Interference (MAI)

 
1
Pb  Q  


K  1 3N   2Eb 


• If the users are assumed to be equal power interferers, can be

analyzed using the central limit theorem (sum of IID RV’s)
Example of Performance Degradation

N=8 N=32
 Near/Far Problem (I)
• Performance estimates derived using assumption that all
users have same power level
• Reverse link (mobile to base) makes this unrealistic since
mobiles are moving
• Adjust power levels constantly to keep equal

k 1
 Near/Far Problem (II)

 
1  1 
Pb Q  
K ( k) (1)
 N   2 E b (1) 
  k2 Eb 3E b


• K interferers, one strong interfering signal dominates

performance
• Can result in capacity losses of 10-30%
Multipath Propagation
 RAKE Receiver

• Received signal sampled at the rate 1/Ts> 2/Tc for detection and
synchronization
• Fed to all M RAKE fingers. Interpolation/decimation unit provides a data
stream on chiprate 1/Tc
• Correlation with the complex conjugate of the spreading sequence and
weighted (maximum-ratio criterion)summation over one symbol